Price Determination in Several International Primary Commodity Markets: A Structural Analysis
Author: E.C. HWA
Publication Date:
01 Jan 1979
eISBN:
9781451956528
Language:
English
Keywords:
SP; money supply; output gap; price change; price equation; price expectation; world commodity market; world price; Commodity prices; Price adjustments; Stocks; price model
Download PDF (2.2 MB)

An important cause of instability in the balance of payments of primary producing countries has been instability in the international prices of primary commodities. Price instability can also cause inflation in the countries that import primary commodities—many of which are industrial countries.1

Abstract

An important cause of instability in the balance of payments of primary producing countries has been instability in the international prices of primary commodities. Price instability can also cause inflation in the countries that import primary commodities—many of which are industrial countries.1

An important cause of instability in the balance of payments of primary producing countries has been instability in the international prices of primary commodities. Price instability can also cause inflation in the countries that import primary commodities—many of which are industrial countries.1

Perhaps because both consuming and producing countries consider price instability undesirable, the drastic changes in prices of primary commodities between 1973 and 1975 attracted much attention and caused particular concern. 2 It is against this background that the United Nations Conference on Trade and Development (UNCTAD) has recently proposed an Integrated Program on Commodities (See UNCTAD (1974).) to stabilize international prices of primary commodities.

The fact that the fluctuations in international prices of primary commodities have been much more pronounced than those in international prices of manufactured goods has often been rationalized on the grounds that international commodity markets are more competitive than the domestic markets for manufactures. This is so because prices are more responsive to shifts in supply and demand in competitive markets than they are in monopolistic or oligopolistic markets. In fact, in the latter, theoretical and empirical evidence has shown that prices are, more often than not, determined by factors that bear little, if any, relationship to short-run variations in demand and supply.

While it has been a well-known proposition even in elementary textbooks of economics that prices in competitive markets respond to demand and supply, economists, as noted by McCallum (1974), have made very little effort to formulate an “operational” version of this proposition so that it could be subjected to an econometric test. McCallum’s attempt to formulate and test a competitive model of price adjustment using data for the U. S. lumber industry is thus a notable exception. In contrast, there have been numerous studies concerned with the formulation and testing of monopolistic models of price dynamics.

In this paper, I will formulate a dynamic disequilibrium model of price adjustments in competitive markets and will subsequently subject this model to an empirical test using the data for seven primary commodities in the UNCTAD proposal. 3 While my attempt here is similar to the McCallum study in spirit, I will formulate and test a competitive model of price adjustment in a structural form. 4 By contrast, the McCallum model was formulated and tested in a reduced form. In addition to being more explicit, a structural model has the advantage over a reduced-form model, in that the former yields structural information about the model that is often buried in the latter.

The purposes of this paper are several. First, it provides a test as to whether the marked fluctuations in the prices of primary commodities traded in the international markets could be well explained by the systematic variations of demand and supply and provides a test of how a competitive model of price adjustments really fares when used to explain the price behavior of primary commodities. Second, it throws light on the price determination mechanism for primary commodities in general, because, as shall be argued later, the empirical literature on competitive price models applied to the international markets of primary commodities has been ambiguous. In particular, the literature has failed to indicate whether international commodity prices are primarily stock-, or flow-, or mixed stock-flow-determined. Finally, it facilitates the analytical assessment regarding the feasibility of stabilizing primary commodity prices through buffer stocks, as is envisaged in the UNCTAD proposal, and/or through other means. For there is no doubt that economic analysis of the likely effects of buffer stocks on commodity prices and export earnings has to rely upon knowledge of the price dynamics in international commodity markets.

The plan of this paper is as follows: Section I gives a brief review of the competitive, structural models of price formation that have been used to analyze international commodity prices. Section II formulates a competitive “stock” disequilibrium model of price adjustments. Section III sets forth the empirical tests of this model, and Section IV gives a summary of conclusions.

I. Review of Structural Models of Competitive Price Adjustment

The specification of a structural equation that describes the price adjustment mechanism in competitive markets for storable commodities (commodities with inventories) has fallen into one of three approaches: prices are flow-, stock-, or stock-flow-determined. 5 Symbolically, they are, respectively:

Pt = f(ΔHt,Zt)(1)
Pt = f(Ht,Zt)(2)
Pt = f(HtCt,HtQt,Zt)(3)

where Pt stands for the world commodity price, Ht for the level of inventory at end of the period, Ct for consumption, Qt for output, and Zt for other variables including the disturbance term. ΔHt is the first difference of Ht, which, by definition, equals Qt–Ct.

The flow approach—equation (1)—assumes that prices are determined by the excess supply (demand) measured by the difference between production and consumption. The stock approach—equation (2)—assumes that the price level is primarily determined by the level of inventory. Its common rationalization has been that this equation reflects a demand equation for stock but has been “normalized” on the price variable. 6 As will be shown later, this specification is far too restrictive, in that it implies that both stock and price are always in equilibrium. Whether a particular commodity market can be regarded as always in equilibrium is, of course, an empirical question, but, in theory, an equilibrium model would not provide an adequate framework to analyze the marked fluctuations in primary commodity prices that may be indications of short-run market disequilibria. The mixed stock-flow approach—equation (3)—is simply a hybrid of the stock and flow approaches.

In view of the fact that these divergent hypotheses have purported to explain an identical phenomenon, that is, the behavior of competitive price adjustments for primary commodities, they perhaps should be interpreted as revealing less about the differences in the market structures than about analysts’ own subjective preferences, which are frequently arrived at in an ad hoc manner. The differences in the specification of the price equation may be traced, in part, to the insufficient efforts that have heretofore been made to study price dynamics in competitive markets. As mentioned earlier, economists have thus far devoted much less effort to the empirical analysis of price dynamics in competitive markets than they have devoted to the analysis of monopolistic markets.

Should price adjustments in the world markets for primary commodities be determined by flow, or stock, or mixed stock-flow disequilibria? Do we really have much leeway in choosing among alternative hypotheses for the purpose of analyzing the behavior of international prices of storable primary commodities? I hope that the following analysis will provide some answers to these questions. In the following section, I will attempt to specify a stock disequilibrium model of price adjustments.

II. Stock Disequilibrium Model of Price Adjustment

For a storable commodity, the following equations are sufficient to define the supply of, and the demand for, stock:

Ct = aPt + bXt + uta<0(4)
Qt = cPt + dYt + vtc>0(5)
Htd = f(Pt+1e−Pt) + gZt + wte>0(6)
Ht = Ht−1 + Qt − Ct(7)

The symbols have these meanings:

Ct = rate of consumption during period t

Qt = rate of production during period t

Htd = demand for stock at the end of period t

Ht = actual stock at the end of period t

Pt = commodity price during period t

Pt+1e = price level expected to prevail at time t+1

ut, vt, wt = disturbance terms in the consumption, production, and demand-for-stock equations, respectively

Xt, Yt, Zt = shifting variables in the consumption, production, and demand-for-stock equations, respectively

Equations (4) and (5) are the familiar demand and supply equations. Equation (6) is a demand equation for stock, the full specification of which will be discussed later.

The price level that balances the short-run demand for, and the short-run supply of, stock can be found by inserting equations (4), (5), and (6) into equation (7).

Ptd = 1c + e + |a|(fPt+1e + bXt − dYt)+ gZt − Ht−1 + ut + wt + vt(8)

Subtracting equation (7) from (6) and then substituting equation (8) into the resulting expression yields

Ptd − Pt = 1c+f+|a|(Htd − Ht)(9)

Equation (9) shows that the deviation of the market price from its desired level (equilibrium) is positively related to the deviation of the actual stock from its desired level (equilibrium) and is negatively related to the absolute sum of short-run elasticities of supply and demand. 7 Further, stock and price must achieve their respective equilibria at the same time.

In the theoretical literature of competitive price adjustments, devices such as auctioneering and recontracting have usually been introduced to enable price to adjust to its equilibrium whenever market disequilibrium exists. 8 I will simply postulate a partial adjustment model for the explanation of price changes

Pt − Pt−1 = r(Ptd−Pt−1)0≤r≤1(10)

where r denotes the parameter measuring the speed of adjustment.

Because the motion of dynamic price adjustments in competitive markets must obey equation (9), this equation must be substituted into (10) in order to derive an internally consistent model. Inserting equation (9) into (10) and collecting terms yields

Pt − Pt−1 = μ(Htd−Ht)(11)

where μ = rk1−r and k = (c + f + |a|)–1.

I now have a dynamic model of competitive price adjustments that stipulates that price changes in a competitive market are a monotonic, nondecreasing function of excess stock demand and, as such, the model is in full agreement with the law of supply and demand as stated in the theoretical literature on price dynamics. While stock disequilibrium acts as a trigger for price changes, the model shows that the actual magnitude of price change for a given amount of excess stock demand depends upon such parameters as the short-run price elasticities of supply and demand, as well as the speed of price adjustments; smaller elasticities and higher speeds of price adjustment are associated with greater price adjustments and vice versa.

In short, the competitive model of price adjustment predicts price changes by combining three factors: (1) the amount of stock disequilibrium,9 (2) the sum of the short-run price elasticities of demand and supply, and (3) the speed of price adjustment in response to market disequilibrium.

Equation (11) cannot be directly tested because desired stock Htd is an unobservable variable. This problem can be solved if, instead, the variables that determine the desired stock are observable. To an identification of these variables I now turn.

First, I consider the transactions demand for stock. To producers, stocks are final products and are held as a buffer against imperfect synchronizations of production and consumption. Producers’ demand for stock can be postulated as a positive function of the rate of consumption. To consumers, stocks are intermediate products or raw materials and are held to ensure smooth production. Consumers’ demand for stock can also be postulated as a positive function of the rate of consumption. Therefore, the rate of consumption of a commodity can be postulated to be a determinant of both producers’ and consumers’ transactions demands for stock.

Another important determinant of desired stock is the expected change of prices (Pt+1e – Pt), where Pt+1e is the price expected for period t + 1. This determinant has been emphasized by the supply-of-storage literature. According to this theory, the desired (optimal) level of inventory of a competitive firm that maximizes its present value is Pt+1e – Pt = F(Htd), where F(Htd) is a marginal storage cost function with F'>0 and F(0)<0. 10 If F(Htd) is linearized, 11 it becomes

Pt+1e − Pt = a0 + a1Htda0<0, a1>0(14)

Finally, I introduce a trend variable T to catch the long-run effect of more efficient inventory control techniques on the desired level of inventory.

In sum, the equation for desired stock may be written as

Htd = aCt + β(Pt+1e − Pt) − γT + ηα,β,γ,η>0(15)

where β = 1a1 and η = −a0a1.

In equation (15), Pt+1e is a variable for which no suitable data can be found. 12 Therefore, an explicit hypothesis on how price expectations in the commodity markets are generated is necessary.

This study will adopt the rational expectations hypothesis of John Muth (1961) in lieu of other well-known hypotheses, such as the adaptive expectations hypothesis or the distributed-lag model in one form or another. Because, in the present analysis, the price itself is an “endogenous” variable, the use of the rational expectations hypothesis will be consistent with the underlying model, one purpose of which is to provide predictions of endogenous variables like price, whereas competing hypotheses do not necessarily guarantee this type of consistency. 13 Following the rational expectations hypothesis, the predicted price for period t + 1 is equal to the expected price predicted by the model, conditional on all information available at the time the prediction is made (time t).

Pt+1e = E(Pt+1/Inft)(16)

where E denotes the expected-value operator and Inft denotes the information set available as of time t. The information set should consist only of the values for the predetermined variables in the model as of time t. A proper selection of the predetermined variables requires knowledge about the “complete” model that has already been outlined by the set of equations (4)–(7). The discussion thus far has focused on the price and the demand-for-stock equations. What remains to be done is to identify those predetermined variables that enter into the equations for the other two endogenous variables in the model—that is, Ct (consumption) and Qt (production)—and to use them as possible candidates in the information set Inft. After this is done, the expected price may be represented by the following linear equation: 14

Pt+1e = α0 + α1PMt + α2Yt + α3Pt−1(16′)

where PMt denotes the world price, Yt denotes world income, and Pt–1 denotes the lagged commodity price.

From equation (16), Pt+1 = Pt+1e + et where et is a stochastic disturbance that has zero mean and is uncorrelated with Pt+1e. Therefore, a simple test of equation (16′) by the ordinary least-squares (OLS) technique can be made by using the actual observed price at the next period, Pt+1, as the dependent variable. The results of this test are reported in Table 1.

Table 1.

Regression Results for the Expected PricePt+1einEquation (16′)1

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The figures in parentheses underneath the respective parameters are t-statistics. The double asterisk (**) denotes significance at the 1 per cent level; the single asterisk (*) denotes significance at the 5 per cent level. Sample periods are as follows: 1955–75 for cocoa, coffee, rubber, and sugar; 1962–75 for cotton; 1956–75 for copper; 1960–75 for tin.

An examination of Table 1 reveals that the world price PM stands out as the most significant variable in almost all regressions; only in the copper equation is it less significant than the world income variable Yt. 15 For this reason and also to mitigate the potential multicollinearity problem, only the world price variable PMt will be used to stand for the price expectations variable Pt+1e. Therefore, equation (16′) is simplified to 16

Pt+1e = α0 + α1PMt(16″)

Inserting equation (16) into (15) yields

Htd = aCt + βα1PMt − βPt − γT + η + βα0(19)

Inserting equation (19) into (11) and collecting terms yields

Pt=b0 + b1Ct + b2PMt − b3T − b4Ht + b5Pt−1(20)
where b0=μ(η+βα0)1+μβb3=μγ1+μβb1=μα1+μβb4=μ1+μβ(21)b2=μβα11+μβb5=11+μβ; μ=rk1−r

Since, according to equation (7), Ht = Ht–1 + Qt – Ct, I can also write equation (20) as

Pt = b0 + (b1 + b4)Ct − b4Qt − b4Ht−1+ b2PMt − b3T + b5Pt−1(20′)

or, after combining Qt and Ht–1, as

Pt = b0 + (b1 + b4)Ct − b4(Qt + Ht−1)+ b2PMt − b3T + b5Pt−1(20″)

When the model is applied to the tin price, two modifications are necessary. First, since Ht measures private stocks, 17 the model needs to be modified to take into account the effect of interventions by both international buffer stock agencies and governments on the level of private inventory. For the commodities studied here, only the international tin market has had an operational buffer stock and has been subjected to significant interventions by the U. S. Government. Therefore, for the private tin stocks, Ht = Ht–1 + Qt – Ct – BFt – Gt where BFt denotes net purchases of stocks by the tin buffer stock, and Gt denotes net purchases of stocks by the United States, during year t.

Second, because of multicollinearity, I have used the “real” price of tin—that is, the price deflated by the index of world prices (PTN/PM)t—as the dependent variable in the regression. A priori, this amounts to constraining the long-run elasticity of the tin price with respect to PMt to have a value of unity. Because my findings generally support the view that the world price has a significant impact on commodity prices, this assumption may not be overly restrictive.

In sum, the model for the tin price becomes either

(P/PM)t = b0 + (b1 + b4)Ct − b4Qt − b4(Ht−1 − BFt − Gt)− b3T + b5Pt−1

or

(P/PM)t = b0 + (b1 + b4)Ct − b4(Qt + Ht−1 − BFt − Gt)− b3T + b5Pt−1

Before turning to the next section, where the empirical tests are done, I will note a number of the model’s features.

First, from a theoretical point of view, the model implies that the flow model of price determination is a special case. For it is clear that, in addition to variables Ct and Qt the level of inventory plays an important role in the price determination process; if the level of inventory turns out to be statistically significant, the flow model of price determination is invalidated. Further, although the model—particularly in the form of equation (20′)—appears to be a mixed stock- and flow-determined model, it is basically “stock-oriented” according to Bushaw and Clower (1954). However, from an empirical point of view, it is still interesting to ask whether the excess flow demand plays an additional role in the otherwise stock-determined price model or, in other words, whether a mixed stock-flow price model of the following type can be supported by the data:

∂Pt∂t = α1(Htd − Ht) + α2(Ct − Qt)α1,α2>0

The answer to this question is no, because the excess flow demand variable (Ct – Qt) is already implicit in the model, as can be seen from either equation (20′) or (20″); the inclusion of the variable (Ct – Qt) adds little, if any, explanatory power to the model.

Second, although the model implies that one minus the coefficient of Pt–1 (i.e., 1 –b5) is not exactly equal to r—the underlying speed-of-adjustment coefficient in the price equation—except when βk = 1, it nevertheless is a positive monotonic function of r. 18 Therefore, a higher value of b5 implies a lower speed of adjustment r and vice versa. In particular, r equals unity if, and only if, b5 = 0; and r equals zero if, and only if, b5 = 1. 19

Third, the model implies that price adjustments caused by market disequilibria are always fast enough to restore the market to short-run equilibrium, when the estimated b5 is not statistically different from zero (or when the implied speed of adjustment is unity). It is only in this case that (1) the observed stock and price would always be located on the intersections between the demand and supply schedules and (2) the demand equation for stock could be obtained simply by inverting the price equation. 20

Fourth, the model implies that the parameters in the equation of the desired level of inventory can be exactly identified. That is, I can derive the parameters in this equation from those in the price equation by using the following set of relationships, which are obtained by solving (21):

α = b1b4β = 1−b5b4βα1= b2b4η + βα0 = b0b4(22)γ = b3b4μ = b4b5

Finally, because the model implies that the estimated coefficients for Qt and Ht–1 should turn out to be the same, it would be particularly interesting to test the model by using equation (20′), which would provide a more stringent test than equation (20″) does.

In the following section, the price model in alternative forms (i.e., equations (20′) and (20″)) will be estimated with the data of seven of the ten core commodities contained in the UNCTAD commodity stabilization program: cocoa, coffee, copper, cotton, rubber, sugar, and tin. 21 The results of the estimations will also be described.

III. Empirical Evidence

In order to test the model, time series on prices, consumption, production, and stocks had to be collected for each commodity. Generally speaking, I found reliable data on prices, consumption, and production—but not on stocks. To circumvent the problem, whenever reliable stock data were unavailable or inconsistent with the data on production and consumption, I estimated them by utilizing the following identity: the change in stocks in every period equals the difference between production and consumption. This procedure requires stock data for at least one year, and the year selected should be the one for which the data on stocks, production, and consumption are in closest agreement. For the index of world prices PMt, I used the United Nations’ export price index of manufactures as a proxy. A detailed description of the sources of data for each variable in the model is given in the Appendix.

Annual time-series data were employed in the estimation. Ideally, one should use much more disaggregated time series, such as quarterly or monthly data, in order to detect properly the dynamic adjustments of prices. Unfortunately, at these levels of disaggregation, reliable data are generally unavailable. However, there is an advantage in using annual data: it allows one to choose a single (spot) price in an important world market as a representative world price, because one year has proven to be sufficient to ensure similar movements of prices in different world commodity markets as a result of competitive arbitrage across markets.

In addition to the OLS approach, the model has also been estimated by the instrumental variable (INV) approach to avoid simultaneous-equation bias. 22 Furthermore, whenever there was strong evidence that the disturbance term was serially correlated, the model was corrected for first-order serial correlation using the Cochrane-Orcutt iteration method. The sample period of estimation is 1955–75, except for cotton and tin, for which the sample periods are, respectively, 1962–75 and 1960–75. The results of OLS estimation are reported in Tables 2 and 3, and the results of INV estimation are reported in Tables 4 and 5.

Table 2.

Regression Results of PriceEquation (20′)1

(Ordinary Least-Squares Estimates)

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The symbol ρ denotes the parameter of the first-order Markov process. The double asterisk (**) indicates that an estimate is significant at the 1 per cent level; the single asterisk (*) indicates that an estimate is significant at the 5 per cent level. The figures in parentheses underneath the respective parameters are t-statistics. Variables with incorrect signs were deleted.

DT = dummy variable. In the coffee equation, Dt represents the impact of the International Coffee Agreement on the coffee price. Dt = 1 in (crop years) 1963/64, 1964/65, 1965/66, 1969/70; Dt = 0.5 in 1966/67. In the tin equation, Dt represents the excessive speculative activities after the 1973 oil embargo. Dt = 1 in 1974; otherwise Dt = 0.

The dependent variable is (P/PM)t. The variable Ht–1 excludes net purchases made by the international tin buffer stock and the United States Government.

Table 3.

Regression Results of Price Equation (20″)1

(Ordinary Least-Squares Estimates)

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The symbol ρ denotes the parameter of the first-order Markov process. The double asterisk (**) indicates that an estimate is significant at the 1 per cent level; the single asterisk (*) indicates that an estimate is significant at the 5 per cent level. The figures in parentheses underneath the respective parameters are t-statistics. Variables with incorrect signs were deleted.

Dt = dummy variable. In the coffee equation, Dt represents the impact of the International Coffee Agreement on the coffee price. Dt = 1 in (crop years) 1963/64, 1964/65, 1965/66, 1969/70; Dt = 0.5 in 1966/67. In the tin equation, Dt represents the excessive speculative activities after the 1973 oil embargo. Dt = 1 in 1974; otherwise Dt = 0.

The dependent variable is (P/PM)T. The variable Ht–1 excludes net purchases made by the international tin buffer stock and the United States Government.

Table 4.

Regression Results of Price Equation(20′)1

(Instrumental Variable Estimates)

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The symbol ρ denotes the parameter of the first-order Markov process. The double asterisk (**) indicates that an estimate is significant at the 1 per cent level; the single asterisk (*) indicates that an estimate is significant at the 5 per cent level. The figures in parentheses underneath the respective parameters are t-statistics. Variables with incorrect signs were deleted.

Dt = dummy variable. In the coffee equation, Dt represents the impact of the International Coffee Agreement on the coffee price. Dt = 1 in (crop years) 1963/64, 1964/65, 1965/66, 1969/70; Dt = 0.5 in 1966/67. In the tin equation, Dt represents the excessive speculative activities after the 1973 oil embargo. Dt = 1 in 1974; otherwise Dt = 0.

The dependent variable is (P/PM)T. The variable Ht–1 excludes the net purchases made by the international tin buffer stock and the United States Government.

Table 5.

Regression Results of Price Equation(20″)1

(Instrumental Variable Estimates)

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The symbol ρ denotes the parameter of the first-order Markov process. The double asterisk (**) indicates that an estimate is significant at the 1 per cent level; the single asterisk (*) indicates that an estimate is significant at the 5 per cent level. The figures in parentheses underneath the respective parameters are t-statistics. Variables with incorrect signs were deleted.

Dt = dummy variable. In the coffee equation, Dt represents the impact of the International Coffee Agreement on the coffee price. Dt = 1 in (crop years) 1963/64, 1964/65, 1965/66, 1969/70; Dt = 0.5 in 1966/67. In the tin equation, Dt represents the excessive speculative activities after the 1973 oil embargo. Dt = 1 in 1974; otherwise Dt = 0.

The dependent variable is (P/PM)T. The variable Ht–1 excludes net purchases made by the international tin buffer stock and the United States Government.

Generally speaking, the results are good; most explanatory variables have the theoretically-expected signs and are significant either at or above a 5 per cent level, according to the t-statistics; the adjusted R2s are usually higher than 0.85. 23 Also, the OLS and INV estimates are not so substantially different as to alter my conclusions in any qualitative way.

Because virtually all the equations contain lagged dependent variables,24 the robustness of regression can better be tested by simulating the model dynamically within the sample, where the lagged dependent variable was replaced in each period by the value predicted by the model in the previous period. The dynamic simulation errors of commodity prices, represented by average absolute errors (AAE) in per cent, are reported in Table 6; and the actual and predicted prices for each commodity are shown in Charts 1–4. 25

Table 6.

Average Absolute Forecasting Error (AAE) and Coefficient of Variation (CV) of Commodity Prices1

article image

The definitions of AAE and CV are as follows:

AAE = Σn|Pi − Ai|/Ain. 100CV = [Σ(Ai − A¯)2n]1/2/A¯

where Pi, Ai, and A denote predicted, actual, and the mean of commodity prices, respectively; n denotes the number of periods simulated. The simulation periods are as follows:

Cocoa: 1955–75

Coffee: 1955–75

Cotton: 1963–75

Rubber: 1955–75

Copper: 1956–75

Tin: 1960–75

Sugar: 1956–75

Chart 1.
Chart 1.

World Cocoa and Coffee Prices, 1955–75

(1970=100)

Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A006

Chart 2.
Chart 2.

World Cotton Prices, 1963–75, and World Rubber Prices, 1955–75

(1970 = 100)

Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A006

Chart 3.
Chart 3.

World Copper and Tin Prices, 1956–75

(1970 = 100)

Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A006

Chart 4.
Chart 4.

World Sugar Prices, 1956–75

(1970 = 100)

Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A006

The errors are generally small, except for sugar, which has the largest simulation error. However, it is not immediately apparent that the competitive model of price adjustment is, indeed, least successful in explaining price movements when the model is applied to sugar, because the sugar price has the largest variation of all seven commodities (See Table 6, column 2.). In order to determine the successfulness of the model when it is applied to commodity prices with different degrees of variation, I adjusted the forecast error by the actual price variation presented by the coefficient of variation; specifically, I divided the forecast error by the coefficient of variation. After this adjustment, the results for sugar no longer stand in such sharp contrast to the results for the other commodities (See Table 6, column 3.).

Although Charts 1–4 reveal that the model tends to smooth out the actual price fluctuations, the model picks up a significant portion of the price variations and the turning points. Thus, the dynamic price model is well confirmed by the data. In addition, I can draw a number of more specific conclusions from the results reported in Tables 2–5.

First, the level of inventory Ht–1 is statistically significant, at either the 5 or the 1 per cent level, for all of the seven primary commodities studied. 26 This result should lead us to reject emphatically the pure flow model of price determination for storable commodities. 27

Second, the model is particularly buttressed by the empirical evidence obtained for cocoa, coffee, cotton, sugar, and tin. This is so because the results have confirmed the theoretical expectation that the independently estimated parameters for Qt and Ht–1 would be nearly identical (See Tables 2 and 4.). This fact is also demonstrated by comparing the standard errors of equation (20′) with those of equation (20″), where the coefficients of Qt and Ht–1 are constrained to be identical. This comparison indicates that the standard errors between the two are not significantly different. 28

Third, the index of world prices PMt is significant in all equations. Because this variable is used as a surrogate for the expected price in the next period, this result suggests that price expectations may play an important role in price formation for primary commodities.

The estimated long-run elasticities of commodity prices with respect to expected prices for cocoa, coffee, cotton, sugar, rubber, and copper are, respectively; 1.2, 0.8, 2.0, 3.8, 0.5, and 1.1. Thus, cotton and sugar prices appear to be more sensitive to the influence of price expectations than the prices of the other five commodities. 29

Fourth, the time trend is significant in the price equations for coffee, cotton, rubber, and tin. This is an indication that the optimal holdings of inventory may have gradually been reduced because more efficient inventory control techniques were used.

Finally, according to the size of the lagged price term Pt–1, the markets for copper and tin appear to adjust more slowly, in response to market disequilibria, than the markets for cocoa, coffee, cotton, sugar, and rubber. Among the agricultural commodities, cotton and sugar appear to have the speediest price adjustment mechanisms, which always ensure market equilibrium within one year. 30

STABILITY TEST

During the period 1973–75, the world economy was characterized by a number of unprecedented events. Among the most significant ones were the following: (1) the change in the international monetary system from a system of par values to one of greater exchange rate flexibility; and (2) the fourfold increase in international oil prices. These events were accompanied by an unprecedented boom in the prices of primary commodities. Thus, it would be interesting to see whether the boom in commodity prices was indeed unusual—that is, to see whether the fundamental relationship determining commodity prices has changed as a result of the boom. 31 In more technical terms, I would like to test the null hypothesis that the price equation estimated by the sample including the period 1973–75 comes from the same population as the price equation estimated by the sample excluding this period. For this test, equation (20″) was re-estimated by the OLS method for a shorter sample period, ending in 1972. The results are reported in Table 7. 32

Table 7.

Regression Results of PriceEquation (20′)1

(Ordinary Least-Squares Estimates)

article image

R2 denotes the coefficient of multiple correlation, adjusted for degrees of freedom; D-W denotes the Durbin-Watson statistic; S. E. denotes the standard error of estimate. The symbol ρ denotes the parameter of the first-order Markov process. The double asterisk (**) indicates that an estimate is significant at the 1 per cent level; the single asterisk (*) indicates that an estimate is significant at the 5 per cent level. The figures in parentheses underneath the respective parameters are t-statistics. The sample periods for all six commodities end in 1972. Variables with incorrect signs were deleted.

Comparing Table 7 with Table 2, we notice a drastic change in the results; the proxy for the expected price PM turns out either to be insignificant or to have the wrong sign, which was not the case in Table 2. This outcome indicates that inflationary expectations might, indeed, have “intensified” during 1973–75, exerting a significant influence on commodity prices. All other variables in the equation, however, maintain the theoretically-expected signs and are usually significant.

The stability of equation (20′) was tested by the familiar Chow test. According to this test, the null hypothesis is rejected for cocoa, copper, cotton, and sugar at the 5 per cent significance level (See Table 8.). However, at the same significance level, the null hypothesis is not rejected for coffee and rubber. Thus, this test suggests that the unusual events that occurred during the period 1973–75 might have had asymmetrical effects on world commodity markets. Given the fact that the greatest change has occurred in the expected price variables, increased price instability may be attributed to changes in price expectations.

Table 8.

Tests of Stability of Price Equation(20′)1

article image

Figures in parentheses under the F-statistics denote degrees of freedom.

IV. Summary of Conclusions

Empirical analysis of competitive price adjustments has rarely been undertaken to date, while numerous efforts have been made to study price dynamics in monopolistic markets. The fact that economists have devoted a great deal of effort to studying price adjustments in monopolistic, rather than in competitive, markets may, to some extent, reflect the fact that competitive markets are rarely found in the sectors of domestic economies that are the usual subjects of analysis. However, it is usually recognized that international prices may be fruitfully analyzed by using models of competitive price adjustment. I believe that international markets for primary commodities could be ideal testing grounds for models of competitive price adjustment.

In this paper, I have formulated a dynamic disequilibrium model of price adjustment in competitive markets. The model has been formulated in the structure-equation form and has been tested using the data for seven of the ten core commodities in the UNCTAD Integrated Program on Commodities. These commodities are: cocoa, coffee, copper, cotton, rubber, sugar, and tin. The major conclusions may be summarized as follows:

(1) The international markets for these primary commodities are consistent with the mode of a competitive market.

(2) The changes in primary commodity prices are determined by stock disequilibria, rather than by either flow or mixed stock-flow disequilibria.

(3) Price expectations had a significant impact on primary commodity prices during the period 1973–75, when world commodity markets witnessed the largest boom-and-bust price cycles since the Second World War. Also, price expectations were significantly influenced by world prices and inflation. Thus, it appears that the world-wide inflation prevailing at that time contributed to the surge in—and, therefore, to the instability of—commodity prices. However, before 1973, price expectations do not seem to have played a significant role in the determination of commodity prices. Thus, the underlying forces that govern price expectations may be very unstable at times. If this is indeed so, it has quite an uncomfortable implication for short-run commodity price forecasting: unless the variability of price expectations can somehow be captured, forecasting commodity prices will be a risky venture.

(4) The estimates of the speed of price adjustment toward short-run equilibrium are found to be generally larger for agricultural commodities than for metals. The estimates further indicate that one year may be sufficient for the agricultural commodity markets to reach equilibrium, while it may not be sufficient for metal markets.

APPENDIX

I. Data Definitions and Sources

(1) Cocoa: Cocoa Statistics (Gill & Duffus Group Limited, London, December 1975).

  • Pt = World price of beans (based on the unit value of cocoa beans imported into the United States), U. S. cents per lb., f.o.b., 1970 = 100

  • Ct = World grindings

  • Qt = World gross crop

  • Ht = World stocks

All quantities are in thousands of metric tons and pertain to crop year (October/September) estimates.

(2) Coffee: Annual Coffee Statistics 1966 (Pan-American Coffee Bureau, New York); International Financial Statistics (International Monetary Fund, Washington), various issues; 1976 Commodity Year Book (Commodity Research Bureau, New York, 1977), pp. 117–19.

  • Pt = Weighted average of the prices for unwashed arabica, robusta, other mild arabica, and Colombian mild arabica. Spot New York, U. S. cents per lb., 1970 = 100

  • Ct = World consumption, computed as the sum of net exports and consumption of the exporting countries (the difference between production and exportable production)

  • Qt = World production

  • Ht = World stocks

All quantities are in thousands of 60 kilo bags and pertain to crop year (October/September) estimates.

(3) Cotton: Cotton-World Statistics (International Cotton Advisory Committee, New York), various issues.

  • Pt = Mexican staple medium (SM) 1-1/32; effective 1960–61, changed to SM 1-1/16; U. S. cents per lb.; 1970 = 100.

  • Ct = World consumption

  • Qt = World production

  • Ht = World stocks

All quantities are in millions of bales and pertain to crop year (August/July) estimates.

(4) Sugar: Sugar Year Book, various volumes, and Statistical Bulletin, various issues (International Sugar Organization, London); International Financial Statistics (International Monetary Fund, Washington), various issues.

  • Pt = Free market price, f.o.b., Caribbean and Brazilian ports. Spot New York, U. S. cents per lb., 1970 = 100

  • Ct = World consumption

  • Qt = World production

  • Ht = World stocks

All quantities are in thousands of metric tons.

(5) Rubber: World Rubber Statistics Handbook, Vol. 1(1946/70), and Rubber Statistical Bulletin, various issues (International Rubber Study Group, London).

  • Pt = Singapore, f.o.b., in bales, No. 1 rolled smoked sheets (RSS), Spot Singapore, Malaysian cents per lb., 1970 = 100

  • Ct = World consumption

  • Qt = World production

  • Ht = World stocks

All quantities are in thousands of metric tons.

(6) Tin: Tin Statistics, 1965–1975; Trade in Tin 1960–1974; and Statistical Bulletin, various issues (International Tin Council, London).

  • Pt = London Market Exchange, cash, standard tin (99.75 per cent), £ per metric ton, 1970 = 100

  • Ct = World consumption

  • Qt = World production

  • Ht = World stocks

  • Gt = Stock purchased by U. S. Government (General Services Administration)

  • BFt = Stock purchased by the International Tin Buffer Stock

All quantities are in thousands of metric tons and pertain to the Western world only.

(7) Copper: Year Book (American Bureau of Metal Statistics, New York), various volumes; and World Metal Statistics (World Bureau of Metal Statistics, London), various issues.

  • Pt = London Market Exchange, £ per metric ton, 1970 = 100

  • Ct = World consumption

  • Qt = World production

  • Ht = World stocks

All quantities are in thousands of metric tons and pertain to the Western world only.

(8) Real World Income (Yt)

The real world income variables used in the regressions of Pt+1e in Table 1 for cocoa, coffee, copper, and tin are expressed in billions of 1970 dollars.

They are the import-share-weighted real gross national products of 14 industrial countries (defined according to International Financial Statistics). The weights are derived from The Significance of Basic Commodities in World Trade in 1970 (United Nations, Document No. A/9544/Add. 1, 1974). The real world income variable (in 1970 prices) used for cotton, rubber, and sugar is an index (1970=100) derived as a weighted average of the real income indices of three groups of economies: developed market economies, developing economies, and centrally planned economies. The data for developed market economies are published by the Organization for Economic Cooperation and Development in its monthly publication, Main Economic Indicators (January 1968); the data for developing and centrally planned economies are published by the United Nations in its 1976 Statistical Yearbook.

REFERENCES

  • Adams, F. G., and Jere R. Behrman, Econometric Models of World Agricultural Commodity Markets: Cocoa, Coffee, Tea, Wool, Cotton, Sugar, Wheat, Rice (Cambridge, Massachusetts, 1976).

    • Search Google Scholar
    • Export Citation

  • Barro, Robert J., “A Theory of Monopolistic Price Adjustment,” Review of Economic Studies, Vol. 39 (January 1972), pp. 17–26.

  • Brennan, Michael J., “The Supply of Storage,” American Economic Review, Vol. 48 (March 1958), pp. 50–72.

  • Bushaw, D. W., and R. W. Clower, “Price Determination in a Stock-Flow Economy,” Econometrica, Vol. 22 (July 1954), pp. 328–43.

  • Cooper, R. N., and R. Z. Lawrence, “The 1972–75 Commodity Boom,” Brookings Papers on Economic Activity: 3 (1975), pp. 671–715.

  • International Monetary Fund, International Financial Statistics, various issues.

  • Kohn, Meir, “Competitive Speculation,” Econometrica, Vol. 46 (September 1978), pp. 1061–76.

  • Labys, Walter C., Dynamic Commodity Models: Specification, Estimation, and Simulation (Lexington, Massachusetts, 1973).

  • Labys, Walter C. and Clive W. J. Granger, Speculation, Hedging, and Commodity Price Forecasts (Lexington, Massachusetts, 1970).

  • McCallum, B. T., “Competitive Price Adjustments: An Empirical Study,” American Economic Review, Vol. 64 (March 1974), pp. 56–65.

  • Muth, John F., “Rational Expectations and the Theory of Price Movements,” Econometrica, Vol. 29 (July 1961), pp. 315–35.

  • Telser, Lester G., “Futures Trading and the Storage of Cotton and Wheat,” Journal of Political Economy, Vol. 66 (June 1958), pp. 233–55.

    • Search Google Scholar
    • Export Citation

  • United Nations Conference on Trade and Development, An Integrated Programme for Commodities, U. N. Document No. TD/B/C. 1/166 (United Nations, Geneva, 1974).

    • Search Google Scholar
    • Export Citation

  • Working, Holbrook, “The Theory of Price of Storage,” American Economic Review, Vol. 39 (December 1949), pp. 1254–62.

SUMMARIES

The “Vicious Circle” Hypothesis—john f.o. bilson (pages 1–37)

Proponents of the vicious circle hypothesis argue that exchange rate depreciation is an independent source of inflationary pressure and an ineffective tool of adjustment policy. In this paper, a general equilibrium model of a small open economy is constructed that reproduces the pattern of price, wage, and exchange rate behavior stressed by the “vicious circle” view in the case where the primary source of disturbances in the economy is monetary. The paper also undertakes a dynamic analysis of the issue of policy effectiveness and relates the effectiveness of monetary and fiscal policy to the speed of price and wage adjustment, the degree of monetary accommodation, and the openness of the economy. Finally, a policy that allows a country to escape from an inflationary spiral without undergoing a prolonged period of unemployment is proposed. The validity of this policy proposal relies heavily on the assumption that capital markets are highly integrated. With integrated capital markets, a policy of fiscal stimulus is shown to lead to an appreciation of the exchange rate, a reduction in the rate of inflation of wages and prices, and a temporary increase in output and employment. This policy, combined with an appropriate degree of monetary constraint, is shown to be capable, in theory, of breaking a wage-price-exchange rate spiral.

Trade, Prices, and Output in Japan: A Simple Monetary Model—bijan b. aghevli and carlos a. rodriguez (pages 38–54)

In order to assess empirically the role of monetary factors in the process of short-run determination of output growth, inflation, and the trade balance, a model is developed for Japan and estimated for the period 1965–76. The model is essentially monetarist, in that the excess supply of cash balances plays the leading role in the short-run adjustment process of the economy. The adjustments of output, prices, and the trade balance are assumed to depend on the excess supply of money, the degree of openness of the economy, and the existing level of excess capacity.

The estimated model tracks the short-run movements of endogenous variables well, notwithstanding the large disturbances associated with the world monetary system, the commodity boom, and the “oil crisis,” all of which occurred during the period. The estimated coefficients indicate that variations in the money supply have a strong impact on the economy. In addition, the level of excess capacity in the economy acts as an important self-equilibrating variable; a larger level of excess capacity is shown to reduce inflation and to stimulate the growth rate of output. Starting with the fairly high level of excess capacity in 1977, the model is simulated to show the reactions of output, prices, and the trade balance to various rates of monetary expansion. At one extreme, a rate of monetary expansion close to 20 per cent is required to bring the economy to its full capacity by 1980. Such an early achievement of full employment, however, would result in a substantial increase in the rate of inflation and would change the present large trade surplus to a deficit within a couple of years.

Fiscal Policy in Oil Exporting Countries, 1972–78—david r. morgan (pages 55–86)

Following the oil price rises of late 1973, several organizations predicted massive accumulations of international reserves in the major oil exporting countries during the remainder of the decade. The predictions have not been realized. Most commentators failed to anticipate the speed with which highly ambitious development strategies could be formulated and implemented in the major oil exporting countries. The growth of government expenditure in these countries over the period 1972–78 was, by any standards, spectacular; in both absolute and percentage terms, it was more rapid than the growth in their revenues. The overall fiscal position of the 12 major oil exporting countries moved from a surplus of almost $40 billion in 1974 to a deficit of approximately $15 billion in 1978. Movements in these countries’ international reserve holdings have been closely correlated with overall fiscal developments.

After examining these developments, the paper shows that the conventional presentation of fiscal and monetary accounts is inappropriate for examining the impact of fiscal policy on the domestic economies of oil exporting countries. An alternative presentation is provided that focuses on the domestic budget deficit and its implications for domestic liquidity creation. Application of the alternative framework to six major oil exporting countries for the period 1972–78 provides strong support for the propositions that the domestic budget deficit is the primary determinant of movements in domestic liquidity and inflation and that fiscal policy must be the primary instrument of demand management. Sharp reductions in the rate of growth of the domestic budget deficit have been closely associated with the restoration of domestic financial stability; earlier attempts to control inflationary pressures through subsidies and price controls, while highly expansionary fiscal policies were pursued at the same time, were not successful.

The paper concludes with a discussion of some issues surrounding the formulation and implementation of appropriate fiscal policies in oil exporting countries. Notwithstanding data deficiencies and rapid structural change, for five oil exporting countries there appears to be a stable relationship between the demand for real liquid balances and real non-oil gross domestic product. This relationship, in conjunction with information concerning the private sector’s balance of payments and domestic bank credit developments, provides an indication of the magnitude of the domestic budget deficit that is consistent with price and real output targets. The primary constraint on the implementation of appropriate fiscal and monetary policies is an inadequate adaptation of economic management infrastructures to the transformation of oil exporting economies.

Measuring the Elasticity of Tax Revenue: A Divisia Index Approach—nurun n. choudhry (pages 87–122)

This paper proposes a method of estimating the elasticity of total tax revenue that does not require the traditional adjustment of historical revenue to eliminate the effects of discretionary tax measures. The proposed method involves three steps. First, the effects of discretionary tax measures on revenue are estimated by an index that separates the automatic growth of revenue from the total growth. Second, the buoyancy of tax revenue is estimated with respect to GDP by a standard regression technique. Finally, the buoyancy estimate obtained by the second step is adjusted, by an appropriate transformation of the index of discretionary revenue as estimated by the first step, in order to provide an estimate of the elasticity of tax revenue.

The proposed index of discretionary revenue is based on the principle of the Divisia index, which is widely used in measuring technical change. This method, however, has two limitations: (1) the Divisia index of discretionary tax change underestimates (overestimates) the positive (negative) revenue effects of such measures; and (2) if discretionary changes produce very large revenue effects, the method gives unsatisfactory results. The main advantage of this method is that it uses only historical data and requires no specific information on the revenue effects or on the frequency of past discretionary tax changes.

The results of the application of this new method of estimating tax elasticity to the United States, the United Kingdom, Malaysia, and Kenya are found to conform to the nature and overall direction of discretionary changes that have occurred in these countries during the periods of investigation.

International Comparisons of Taxation for Selected Developing Countries, 1972–76—alan a. tait, wilfrid l.m. grätz, and barry j. eichengreen (pages 123–56)

The paper reviews the controversial measures of “tax effort,” updates previous studies, and compares current results to earlier findings. Some trends in taxation are described. A new, broader sample of countries is used to show the vulnerability of absolute tax indices to changes in the sample. However, rankings prove to be relatively stable.

Price Determination in Several International Primary Commodity Markets: A Structural Analysis—e.c. hwa (pages 157–88)

Empirical analysis of competitive price adjustments has rarely been undertaken to date, while numerous studies have attempted to analyze price dynamics in monopolistic markets. In this paper, I have formulated a dynamic disequilibrium model of price adjustments in competitive markets. The model has been formulated in the structure-equation form and tested by using data for seven of the ten “core” commodities contained in the United Nations Conference on Trade and Development (UNCTAD) Integrated Program on Commodities. These commodities are cocoa, coffee, cotton, sugar, rubber, copper, and tin. The major conclusions may be summarized as follows: (1) The international markets for these primary commodities are consistent with the mode of a competitive market; (2) The changes in primary commodity prices are determined by “stock” disequilibria rather than by either “flow” or mixed “stock-flow” disequilibria; (3) Price expectations had a significant impact on primary commodity prices during the period 1973–75, when world commodity markets witnessed the largest boom-bust price cycles since the Second World War. Also, they were significantly influenced by the high world price level. Thus, it appears that the world-wide inflationary condition then prevailing contributed to the surge in—and, therefore, to the instability of—commodity prices. However, before 1973, price expectations do not seem to have played a significant role in the determination of commodity prices. Thus, the underlying forces that govern price expectations may at times be very unstable; and (4) The speed of price adjustments toward short-run equilibrium is generally faster for the agricultural commodities than for the metals included in this study. The estimates further indicate that one year may be sufficient for agricultural commodity markets to reach equilibrium but may not be sufficient for metal markets.

RESUMES

L’hypothèse du “cercle vicieux”—john f.o. bilson (pages 1–37)

Les défenseurs de l’hypothèse du cercle vicieux soutiennent que la dévaluation d’une monnaie est à la fois une source indépendante de pressions inflationnistes et un instrument inefficace de politique d’ajustement. Dans cette étude, l’auteur présente un modèle d’équilibre général pour un petit pays à économie ouverte, qui reproduit le schéma de comportement des prix, des salaires et du taux de change mis en relief par la thèse du “cercle vicieux”, dans le cas où la principale source de perturbations dans l’économie est d’ordre monétaire. Dans le meme article, l’auteur procède également à une analyse dynamique de la question de l’efficacité de la politique économique et établit un rapport entre l’efficacité de la politique monétaire et budgétaire et la vitesse d’ajustement des prix et des salaires, la situation de l’offre et de la demande de monnaie et l’ouverture de l’économie. Pour finir, l’auteur propose une politique qui permet aux pays d’échapper à la spirale inflationniste sans qu’ils aient à enregistrer une période de chômage prolongée. La validité de cette proposition repose dans une très large mesure sur l’hypothèse que les marchés de capitaux sont fortement intégrés. Avec des marchés de capitaux intégrés, il apparaît qu’une politique de stimulation budgétaire entraîne une appréciation du taux de change, une réduction du taux d’augmentation des salaires et des prix et un accroissement temporaire de la production et de l’emploi. L’auteur montre qu’une telle politique, assortie d’un certain degré d’austérité monétaire, est théoriquement capable d’arrêter la spirale salaires-prix-taux de change.

Commerce, prix et production au Japon: un modèle monétaire simple—bijan b. aghevli et carlos a. rodriguez (pages 38–54)

La présente étude constitue une tentative d’évaluation empirique du rôle des facteurs monétaires dans le processus de détermination à court terme de la croissance de la production, de l’inflation et de la balance commerciale au Japon entre 1965 et 1976. Le modèle testé est essentiellement monétariste; en effet, l’excédent d’encaisses monétaires joue un rôle prépondérant dans le processus d’ajustement à court terme de l’économie. Par hypothèse, les ajustements de la production, des prix et de la balance commerciale dépendent de l’offre excédentaire de monnaie, du degré d’ouverture de l’économie et du niveau existant de capacité de production excédentaire.

Le modèle estimé retrace bien les mouvements à court terme des variables endogènes, malgré les perturbations importantes liées au système monétaire mondial, à la forte hausse des prix des produits de base et à la crise pétrolière, qui sont survenues pendant cette période. Les coefficients estimés indiquent que les variations de la masse monétaire exercent une action stabilisatrice considérable. En outre, le niveau de capacité excédentaire dans l’économie agit comme une variable importante qui s’équilibre d’elle-même; il apparait qu’un niveau plus élevé de capacité excédentaire réduit l’inflation et stimule le taux de croissance de la production. A partir du niveau relativement élevé de capacité excédentaire de 1977, une simulation est effectuée pour montrer les réactions de la production, des prix et de la balance commerciale à divers taux d’expansion monétaire. A l’un des extrêmes, un taux d’expansion monétaire de près de 20 pour 100 est nécessaire pour que l’écart de production entre production effective et production potentielle soit comblé totalement en 1980. Toutefois, la réalisation aussi précoce du plein emploi entraînerait une augmentation très sensible du taux d’inflation et transformerait le fort excédent commercial actuel en déficit, en l’espace de deux ans.

La politique budgétaire des pays exportateurs de pétrole, 1972–78—david r. morgan (pages 55–86)

A la suite de la hausse des prix pétroliers à la fin de 1973, plusieurs organisations avaient prédit pour le reste de la décennie une accumulation massive de réserves internationales dans les principaux pays exportateurs de pétrole. Ces prédictions ne se sont pas réalisées. La plupart des observateurs n’avaient pas prévu la rapidité avec laquelle des stratégies de développement très ambitieuses pourraient être formulées et mises en oeuvre dans ces pays. Entre 1972 et 1978, l’accroissement des dépenses publiques y a été incontestablement spectaculaire; il a été plus rapide que celui des recettes, aussi bien en valeur absolue qu’en pourcentage. La situation budgétaire globale des douze principaux pays exportateurs de pétrole s’est renversée, passant d’un excédent de près de 40 milliards de dollars en 1974 à un déficit d’environ 15 milliards de dollars en 1978. Les mouvements des réserves internationales de ces pays ont été étroitement liés à l’évolution de leur situation budgétaire globale.

Après avoir examiné cette évolution, l’auteur de l’article montre que la présentation classique des comptes budgétaires et monétaires ne convient pas pour étudier l’impact de la politique budgétaire sur l’économie intérieure des pays exportateurs de pétrole. David Morgan propose une présentation différente qui met l’accent sur le déficit budgétaire d’origine interne et son incidence sur la création de liquidité intérieure. L’application de ce cadre d’analyse à six des principaux pays exportateurs de pétrole pour la période 1972–78 tend à confirmer que le déficit budgétaire d’origine interne constitue le principal facteur déterminant les variations de la liquidité et du taux d’inflation intérieur et que la politique budgétaire doit être l’instrument primordial de gestion de la demande. De fortes réductions du taux d’accroissement du déficit budgétaire d’origine interne ont été étroitement liées au rétablissement de la stabilité financière intérieure; les pays qui ont tenté dans le passé de contenir les pressions inflationnistes au moyen de subventions et de contrôles des prix tout en poursuivant une politique budgétaire fortement expansionniste ont échoué.

L’article se termine par une discussion de certains problèmes relatifs à la formulation et à la mise en oeuvre de politiques budgétaires appropriées dans les pays exportateurs de pétrole. En dépit de certaines déficiences en ce qui concerne les données et de changements structurels rapides, il semble que dans cinq pays exportateurs de pétrole une relation stable existe entre la demande d’encaisses réelles et le produit intérieur brut réel non pétrolier. Cette relation ainsi que les données de la balance des paiements du secteur privé et l’évolution du crédit bancaire intérieur donnent une indication du niveau auquel le déficit budgétaire d’origine interne est compatible avec les objectifs de prix et de production réelle. La principale contrainte dans la mise en oeuvre de politiques budgétaire et monétaire appropriées réside dans la mauvaise adaptation des infrastructures de la gestion économique à la transformation de l’économie des pays exportateurs de pétrole.

Mesure de l’élasticité automatique des recettes fiscales: une méthode dérivée de l’indice Divisia—nurun n. choudhry (pages 87–122)

Dans le présent document l’auteur propose une méthode d’estimation de l’élasticité automatique des recettes fiscales totales qui évite au statisticien d’avoir à ajuster—comme c’est le cas dans les autres méthodes—les séries chronologiques des recettes pour éliminer les effets des mesures fiscales discrétionnaires. La méthode en question se décompose en trois étapes: premièrement, on estime les effets des mesures fiscales discrétionnaires sur les recettes à l’aide d’un indice qui sépare la croissance automatique des recettes de la croissance totale. Deuxièmement, on estime l’élasticité globale des recettes fiscales par rapport au PIB par une simple analyse de régression. Enfin, on ajuste l’estimation de l’élasticité globale obtenue par le calcul ci-dessus en procédant à une transformation appropriée de l’indice des recettes provenant des mesures discrétionnaires, estimé lors de la première étape des calculs, pour arriver à une estimation de l’élasticité automatique des recettes fiscales.

L’indice des “recettes discrétionnaires” proposé dans le présent document est fondé sur le principe de l’indice Divisia, qui est d’une utilisation très répandue pour mesurer les changes techniques. Cette méthode comporte cependant deux limitations: 1) l’indice Divisia des variations des recettes fiscales résultant des changements apportés à la législation fiscale sous-estime (surestime) les effets positifs (négatifs) de ces mesures sur les recettes; et 2) au cas où des changements discrétionnaires ont des effets très sensibles sur les recettes, la méthode donne des résultats peu satisfaisants. Le principal avantage de cette méthode tient au fait qu’elle n’utilise que des séries chronologiques et n’exige pas que l’on dispose d’informations spécifiques sur les effets sur les recettes ni sur la fréquence des modifications apportées dans le passé à la législation fiscale.

Appliquée dans le cas des Etats-Unis, du Royaume-Uni, de la Malaisie et du Kenya, cette méthode d’estimation de l’élasticité automatique des recettes fiscales donne des résultats qui apparaissent conformes à la nature et à l’orientation générale des mesures discrétionnaires prises dans ces pays au cours des périodes couvertes par l’enquête statistique.

Comparaisons entre les systèmes fiscaux de certains pays en développement, 1972–76—alan a. tait, wilfrid l.m. grätz et barry J. eichengreen (pages 123–56)

Dans cette étude, les auteurs examinent la manière controversée de mesurer l’“effort fiscal”, mettent àjour les études antérieures et comparent les résultats actueis aux résultats antérieurs. Les auteurs y décrivent certaines tendances de la fiscalité. Ils se servent d’un nouvel échantillon comprenant un plus grand nombre de pays pour montrer à quel point les indices fiscaux absolus sont sensibles aux variations de l’échantillon. Les classements établis s’avèrent, toutefois, relativement stables.

La détermination des prix sur plusieurs marchés internationaux de produits primaires de base: Analyse structurelle—e.c. hwa (pages 157–88)

On constate que, jusqu’à present, il n’existe que peu d’études empiriques des ajustements de prix sur les marchés concurrentiels, alors qu’au contraire de nombreuses études ont été consacrées à la dynamique des prix sur les marchés monopolistiques. Dans le présent document, l’auteur a construit un modèle de déséquilibre dynamique permettant d’analyser l’évolution des prix sur les marchés concurrentiels. Le modèle a été formulé sous une forme structurelle et testé à partir des statistiques obtenues pour sept des dix produits “clés” inclus dans le programme intégré pour les produits de base adopté par la Conférence des Nations unies sur le commerce et le développement. Ces produits sont le cacao, le café, le coton, le sucre, le caoutchouc, le cuivre et l’étain. Les principales conclusions de l’étude peuvent se résumer comme suit: 1) les marches internationaux pour ces produits de base s’apparentent, par leur nature, aux marchés concurrentiels; 2) les variations des cours des produits de base sont déterminées par des déséquilibres de “stocks”, plutôt que par soit des déséquilibres de “flux”, soit des déséquilibres des deux; 3) les expectations de prix ont eu une incidence significative sur les cours des produits de base pendant la période 1973–75, époque pendant laquelle les marchés mondiaux des produits de base ont enregistré les cycles les plus amples de hausse et de baisse des cours depuis la seconde guerre mondiale. Elles ont été aussi influencées sensiblement par le niveau élevé des cours mondiaux. Il semble donc que la situation inflationniste mondiale qui prévalait alors a contribué à la montée en flèche—et, par conséquent, à l’instabilité—des cours des produits de base. Avant 1973, toutefois, les expectations de prix n’ont pas joué, apparemment, de rôle significatif dans la détermination des cours des produits de base. D’où l’on peut déduire que les forces qui sont à l’origine des expectations de prix peuvent être parfois très instables; enfin 4) l’ajustement des prix vers un équilibre à court terme est, en général, plus rapide pour les produits agricoles que pour les métaux couverts par la présente étude. Il ressort en outre des estimations qu’une période d’un an peut suffire pour que les marchés des produits de base agricoles atteignent l’équilibre, alors qu’elle peut être insuffisante pour les marchés des métaux.

RESUMENES

La hipótesis del “círculo vicioso”—john f.o. bilson (páginas 1–37)

Quienes propugnan la hipótesis del “círculo vicioso” mantienen que la depreciación del tipo de cambio constituye una fuente independiente de presiones inflacionistas y un instrumento ineficaz de la política de ajuste. En el presente trabajo, se construye un modelo de equilibrio general de una pequeña economía abierta, en el que se reproduce el comportamiento de los precios, salarios y tipo de cambio presentado en la mencionada hipótesis, en el caso de que las principales perturbaciones de la economía sean de origen monetario. Se efectúa también en el presente trabajo un análisis dinámico de la cuestión de la eficacia de las medidas de política y se relaciona la eficacia de la política fiscal y monetaria con la rapidez del ajuste de los precios y salarios, el grado de adaptación monetaria y el grado de apertura de la economía. Por último, se propone una política mediante la cual el país pueda zafarse de una espiral inflacionaria sin sufrir un período prolongado de desempleo. La validez de esta propuesta de política descansa principalmente en el supuesto de que los mercados de capital estén muy integrados. Con mercados de capital integrados, se demuestra que una política de estímulo fiscal ocasiona una apreciación del tipo de cambio, una reducción de la tasa de inflación de precios y salarios y un aumento temporal de la producción y el empleo. Se demuestra también que, en teoría, esta política, en combinación con un grado apropiado de contención monetaria, es capaz de romper la espiral salarios-precios-tipo de cambio.

Comercio, precios y producto del Japón: Un modelo monetario sencillo—bijan b. aghevli y carlos a. rodriguez (Páginas 38–54)

En el presente estudio se pretende evaluar empíricamente la función de los factores monetarios en la determinación a corto plazo del aumento del producto, la inflación y la balanza comercial de Japón en el período 1965–76. El modelo que se somete a prueba es esencialmente monetarista, ya que el exceso de oferta de saldos de efectivo desempeña la función principal en el proceso de ajuste a corto plazo de la economía. Se parte del supuesto de que el ajuste del producto, los precios y la balanza comercial dependen del exceso de oferta monetaria, el grado de apertura de la economía y el nivel existente de capacidad sin utilizar.

En el modelo estimado se siguen bien las variaciones a corto plazo de las variables endógenas, no obstante las grandes perturbaciones relacionadas con el sistema monetario mundial, la subida de los precios de los productos primarios y la “crisis del petróleo”, todo lo cual ocurrió durante el período. Los coeficientes estimados indican que las variaciones de la oferta monetaria surten efectos de estabilización significativos. Además, el grado de capacidad sin utilizar en la economía sirve de importante variable autoequilibradora; se hace ver que el aumento de la capacidad sin utilizar reduce la inflación y estimula el ritmo de crecimiento del producto. Partiendo del nivel bastante alto de exceso de capacidad en 1977, se simulan en el modelo las reacciones del producto, los precios y la balanza comercial con respecto a varias tasas de expansión monetaria. En el punto máximo, hace falta una tasa de expansión monetaria próxima al 20 por ciento para que, en 1980, se cierre por completo la brecha del producto. Sin embargo, tan pronta consecución del pleno empleo produciría un considerable aumento de la tasa de inflación y, en un par de años, convertiría en déficit el cuantioso superávit comercial de la actualidad.

La política fiscal de los países exportadores de petróleo, 1972–78—david r. morgan (páginas 55–86)

A raíz del alza del petróleo producida a fines de 1973, varias organizaciones predijeron que durante el resto del decenio los principales países exportadores de petróleo acumularían cuantiosas reservas internacionales. Esas predicciones no se han cumplido. La mayoría de los observadores no previeron la celeridad con que podrían formularse e implantarse las muy ambiciosas estrategias de desarrollo de los principales países exportadores de petroleo. En el período 1972–78, el gasto público de esos países acusó un crecimiento espectacular, cualquiera sea el punto de comparación con que se lo mida; creció, en términos absolutos y porcentuales, a un ritmo mayor que el ingreso. La posición fiscal agregada de los doce principales países exportadores de petróleo paso de un superávit de aproximadamente $40.000 millones en 1974 a un déficit de unos $15.000 millones en 1978. Las variaciones de las tenencias de reservas internacionales de esos países han guardado una relación muy estrecha con la evolución fiscal global.

Tras el examen de esa evolución, se señala en este trabajo que la presentación convencional de las cuentas monetarias y fiscales no sirve para analizar el impacto dejado por la política fiscal en la economía nacional de los países exportadores de petróleo. Como alternativa, se ofrece una presentación que hace hincapié en el déficit presupuestario interno y sus repercusiones en la creación de liquidez interna. La aplicación del nuevo enfoque a seis importantes países exportadores de petróleo, durante el período 1972–78, corrobora en gran medida el supuesto de que el déficit presupuestario interno es el factor primordial en las variaciones de la liquidez interna y la inflación y que la política fiscal debe ser el instrumento principal de gestión de la demanda. La pronunciada disminución de la tasa de crecimiento del déficit presupuestario interno ha estado estrechamente ligada al restablecimiento de la estabilidad financiera interna, mientras que los intentos anteriores destinados a controlar las presiones inflacionarias mediante subsidios y control de precios, aplicando simultáneamente políticas fiscales muy expansionistas, no han dado fruto.

El trabajo finaliza con el análisis de algunos temas en torno de la formulación e implantación de políticas fiscales adecuadas en los países exportadores de petróleo. Pese a la deficiencia de los datos y a la rapidez de los cambios estructurales, puede afirmarse que en cinco de los países exportadores de petróleo parece haber una relación estable entre la demanda de saldos líquidos reales y el producto interno bruto real no petrolero. Esta relación, junto con la información atinente a la balanza de pagos del sector privado y la evolución del crédito bancario interno, da la pauta del déficit presupuestario interno que es compatible con los objetivos de precio y producto real. El principal condicionamiento que presenta la implantación de políticas fiscales y monetarias adecuadas reside en la insuficiente adaptación que la infraestructura de la gestión económica ofrece ante el proceso de transformación de las economías de los países exportadores de petróleo.

Medición de la elasticidad del ingreso tributario: Método basado en el índice de Divisia—nurun n. choudhry (páginas 87–122)

En este estudio se propone un método de estimación de la elasticidad del ingreso tributario total que no necesita el ajuste tradicional del ingreso tributario histórico para eliminar los efectos de las medidas tributarias discrecionales. El método propuesto consta de tres etapas. En primer lugar, se estiman los efectos de las medidas tributarias discrecionales en el ingreso tributario mediante un índice que permite separar el crecimiento automático del ingreso de su crecimiento total. En segundo lugar, se estima la capacidad de reacción del ingreso tributario con respecto al PIB, mediante una técnica de regresión estándar. Finalmente, se ajusta la estimación de la capacidad de reacción obtenida en la segunda etapa, mediante la transformación apropiada del índice del ingreso resultante de medidas discrecionales estimado en la primera etapa, a fin de obtener una estimación de la elasticidad del ingreso tributario.

El índice propuesto para medir el efecto de las medidas discrecionales se basa en el principio del índice de Divisia, cuyo uso está muy difundido para medir el cambio tecnológico. No obstante, este método tiene dos limitaciones: 1) el índice de Divisia de las modificaciones discrecionales de los impuestos subestima (sobreestima) los efectos positivos (negativos) de dichas medidas en el ingreso tributario, y 2) si las modificaciones discrecionales producen efectos de gran magnitud en el ingreso tributario, los resultados de este método son insatisfactorios. La principal ventaja de este método consiste en que emplea solamente datos históricos y no requiere información específica alguna sobre los efectos de las modificaciones tributarias discrecionales pasadas en el ingreso o sobre la frecuencia de las mismas.

Los resultados de la aplicación de este nuevo método de estimar la elasticidad tributaria en el caso de Estados Unidos, el Reino Unido, Malasia y Kenya están en consonancia con el carácter y la orientación general de las modificaciones discrecionales que han tenido lugar en estos países durante los períodos comprendidos en la investigación.

Comparaciones internacionales de tributación entre determinados países en desarrollo, 1972–76—alan a. tait, wilfrid l.m. grätz y barry j. eichengreen (páginas 123–56)

En este estudio se analizan las discutidas formas de medir el “esfuerzo fiscal”, se ponen al día otros estudios anteriores y se comparan los resultados actuales con los anteriormente hallados. Se describen algunas tendencias de la tributación. Se utiliza una muestra nueva y más amplia de países con el fin de comprobar la vulnerabilidad de los índices tributarios absolutos a las variaciones de la muestra. Con todo, el orden de los países en los grupos resultó ser bastante estable.

Determinación del precio en varios mercados internacionales de productos primarios: Un análisis estructural—e. c. hwa (páginas 157–88)

Hasta la fecha, raramente se han efectuado análisis empíricos de los ajustes competitivos de precios, habiéndose efectuado en cambio numerosos estudios para analizar la dinámica de los precios en mercados monopolísticos. En el presente estudio, he formulado un modelo de desequilibrio dinámico de los ajustes de precios en mercados competitivos. Se ha formulado el modelo en forma de ecuación estructural y se le ha sometido a prueba utilizando datos para siete de los diez productos centrales que componen el Programa Integrado para los Productos Básicos, de la Conferencia de las Naciones Unidas sobre Comercio y Desarrollo (UNCTAD). Dichos productos son: cacao, café, algodón, azúcar, caucho, cobre y estaño. Las principales conclusiones pueden resumirse de la forma siguiente: 1) los mercados internacionales de los mencionados productos primarios concuerdan con la pauta de un mercado competitivo, 2) las variaciones de los precios de los productos primarios vienen determinadas por desequilibrios de “masa”, y no por desequilibrios de “flujos” o de “masa y flujos”, 3) las expectativas de precios ejercieron un impacto significativo en los precios de los productos primarios en 1973–75, período en el que los mercados mundiales de productos primarios experimentaron los mayores altibajos de precios desde la segunda guerra mundial. También influyó en ellos de forma significativa el alto nivel de los precios mundiales. Así pues, parece que la situación inflacionaria existente a nivel mundial contribuyó al alza—y, por consiguiente, a la inestabilidad—de los precios de los productos primarios. Sin embargo, las expectativas de precios no parecen haber desempeñado, con anterioridad a 1973, un papel significativo en la determinación de los precios de los mencionados productos. Por lo tanto, las fuerzas fundamentales que alteran las expectativas de precios pueden ser a veces muy inestables, y 4) los ajustes de precios hacia el equilibrio a corto plazo se efectúan en general más rápidamente en el caso de los productos agrícolas que en el de los metales incluidos en el presente estudio. Más concretamente, las estimaciones indican que en los mercados de productos agrícolas un año puede ser suficiente para alcanzar el equilibrio, pero no así en los mercados de metales.

In statistical matter (except in the resumes and resúmenes) throughout this issue,

Dots (…) indicate that data are not available;

A dash (—) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist;

A single dot (.) indicates decimals;

A comma (,) separates thousands and millions;

“Billion” means a thousand million;

A short dash (–) is used between years or months (e.g., 1975–78 or January-October) to indicate a total of the years or months inclusive of the beginning and ending years or months;

A stroke (/) is used between years (e.g., 1977/78) to indicate a fiscal year or a crop year;

Components of tables may not add to totals shown because of rounding.

International Monetary Fund, Washington, D.C. 20431 U.S.A.

Telephone number: 202 477 7000

Cable address: Interfund

LEGAL AND INSTITUTIONAL ASPECTS OF THE INTERNATIONAL MONETARY SYSTEM: SELECTED ESSAYS

Joseph Gold

This volume reproduces, with slight revisions, 14 essays originally contributed by the author to books and periodicals published under auspices other than the Fund. Attached to each essay is a note that comments briefly on the relationship of the Second Amendment of the Fund’s Articles, which became effective on April 1, 1978, to the subject matter of the essay.

The themes of the essays are linked together by a new introductory essay. The collection as a whole can be regarded as a resumption of the author’s discussion of constitutional development and change in the Fund, which he contributed to Volume II of the Fund’s History of its first twenty years. A major theme is the need for an international monetary system that is regulated by international law, with the Fund at its center. Another major theme is that the legal and institutional aspects of the system should provide flexibility in relation to evolution of the system, its day to day operation, and the handling of crises.

The subjects of the essays include various facets of reform of the international monetary system, the negotiation of change, techniques of flexibility, “sanctions,” collaboration as a source of law, and creation of the SDR.

The development of policies of conditionality in connection with the use of the Fund’s resources, the metamorphosis of the Fund’s exchange transactions, and the invention and development of the stand-by arrangement are discussed from the standpoint of flexibility. The Fund’s principle of uniformity and the former par value system are discussed from the standpoint of inflexibility. Annexed to the essay on the par value system are the legal texts of the Fund on exchange arrangements from the origin of the Fund to the present day.

Pp. xx+633 Price: $17.50

About the Author

Mr. Gold is a graduate of the Law Schools of the Universities of London and Harvard. He has been a member of the staff of the Legal Department of the Fund since 1946, and has been the General Counsel of the Fund as well as the Director of the Legal Department since 1960.

For information and to place orders, write to The Secretary, International Monetary Fund, Washington, D.C. 20431 U.S.A.

*

Mr. Hwa, economist in the Commodities Division of the Research Department, holds degrees from Cornell University. He has held research positions at Cornell University’s Center for Quantitative and Mathematical Research in Economics and Management, at the National Bureau of Economic Research, and at the Stanford Research Institute.

In addition to colleagues in the Fund, the author would like to thank P. Schelde Anderson, Sarra Chernick, Gary Fromm, David Crary, and Walter Labys for their very helpful comments.

1

It is sometimes argued that boom-and-bust cycles in primary commodity prices cannot be the underlying cause of general inflation, for they can affect only relative prices—not all money prices. But there are situations in which these cycles can have inflationary consequences, such as when the ratchet effect is at work.

2

The aggregate spot export price index (1970=100) of primary commodities compiled by the International Monetary Fund reached 212.1 in 1974—its highest level since 1957—and then dropped to 174.1 in 1975 (See the Fund’s monthly publication, International Financial Statistics.). The annual percentage changes in this index during 1973, 1974, and 1975 were, respectively, 54.7, 27.8, and -17.9. The aggregate index masks the fluctuations in its constituent commodity prices, which often display much greater variation.

3

The seven commodities are cocoa, coffee, sugar, rubber, cotton, tin, and copper.

4

The model is structural in the sense that, as will be clear later, consumption and production are not substituted for their respective underlying determinants.

5

See Walter C. Labys (1973, pp. 85–105). In this study, I will only be concerned with storable commodities.

6

See, for instance, Adams and Behrman (1976).

7

If equations (4)–(6) are log-linear functions, then c, f, and a are elasticities.

8

Barro (1972) has taken a different view. He argues that the response of prices to disequilibrium is essentially a monopolistic phenomenon.

9

If flow disequilibrium is defined as the gap between imports and exports, rather than between consumption and production, flow disequilibrium is equivalent to stock disequilibrium. The proof is given here. I write two equations defining, respectively, aggregate exports and aggregate imports of a commodity. And, without loss of generality, I assume that exports and imports are, respectively, net exports and net imports. Thus Et = Qxt − Cxt − (Hxtd − Hxt−1)(12) Mt = Cmt − Qmt + (Hmtd − Hmt−1)(13) where Qxt, Cxt, Hxtd and Hxt denote, respectively, production, consumption, demand for stock, and actual stock of the exporting countries as a whole; Qmt, Cmt, Hmtd, and Hmt denote the corresponding variables of the importing countries, taken as a whole. Subtracting equation (12) from (13), I obtain Mt − Et=(Hxtd + Hmtd) − (Qxt + Qmt)− (Cxt + Cmt) + (Hxt−1 + Hmt−1)=Htd − (Qt −Ct + Ht−1)= Htd − Ht

10

See Working (1949), Telser (1958), and Brennan (1958). Meir Kohn (1978) has shown that such a relationship can also be derived from the rational behavior of individual speculators. Thus, Pt+1e – Pt could also pick up some of the speculative demand for stock.

11

This assumption was also made by McCallum (1974).

12

It seems that a natural candidate for Pt+1e is the futures price. However, the findings of Labys and Granger (1970) indicate that the futures price does not seem to lead the spot price and, therefore, does not seem to be a useful predictor. The implication is that the futures price primarily reflects financing costs and other costs of carrying inventory, but not the expected spot price.

13

For instance, there is no reason for the price predictions generated by a distributed-lag hypothesis to be consistent with those implied by the underlying model. However, the argument is not meant to imply that the rational expectations hypothesis is free from any defect. Indeed, the rational expectations hypothesis might give market participants more information than they actually possess.

14

The world consumption function of a commodity is postulated to be a function of its own price and world money income, Ct = φ1(Pt,PMtYt),∂Ct∂Pt<0, ∂Ct∂PMt>0,∂Ct∂Yt>0(17) where Pt denotes the “own” price, PMt denotes the world price, and Yt denotes real world income (PMt Yt, thus, is world money income.). This equation may be viewed as having been derived from the optimizing behavior of consumers under the budget constraint. The exact functional form of the consumption function depends on the specification of an explicit utility function, which is not specified here because this is not necessary for the present purpose. The function postulated in equation (17), however, reflects a rather restrictive utility function, in that consumption does not depend on the prices of substitutes.

Similarly, the level of output that maximizes a short-run profit function, subject to the constraint of a short-run production function under perfectly competitive conditions of both product and factor markets, can be shown to be Qt = φ2(Pt−1, T) ∂Qt∂Pt−1>0,∂Q∂T>0(18) where T denotes a time trend, standing for the productive capacity or capital stock in the short-run production function. The short-run variable costs, such as labor and raw materials costs, however, are ignored in equation (18).

Equations (7), (15), (17), and (18) form a simultaneous-equation model that determines four endogenous variables—Ct, Qt, Pt, and Ht. The predetermined variables are PMt, Yt, T, and Pt–1. However, the time trend is excluded from the list of predetermined variables in equation (16′) because it is highly correlated with the real world income variable.

15

In the equations for cotton and copper, PMt becomes highly significant after Yt and T are deleted.

Using a 5 per cent confidence interval, the Durbin-Watson statistics reveal that the hypothesis of first-order serial correlation should be rejected for all commodities, except for cocoa.

16

This simplification will reduce somewhat the predictive power that accrues to the market participants, but the reduction will probably not be considerable. It will affect the overall predictive power of the price equation even less, because Yt is likely to be collinear with both Ct and PMt, and Pt–1 is already in the price equation.

17

Private stocks are stocks held by private traders; they exclude those stocks held by national authorities and international buffer stock agencies.

18

Because b5 = 11 + μβ and μ = rk1 − r, 1−b5 = rkβ1−r(1−kβ). Therefore, ∂r∂(1−b5) = βk[1−r(1−βk)]2, which is positive because kβ>0.

20

This can be easily seen by comparing equations (19) and (20) when b5=0.

21

The other three core commodities are jute, sisal, and tea.

22

The instruments used for each commodity are world income, the world price, and lagged commodity prices.

23

The variables with incorrect signs were deleted, and the regression was rerun. However, the variables that had correct signs were retained in the regression, even if they were statistically insignificant.

24

Note that the equations for cotton and sugar were corrected for first-order serial correlation, so that the lagged dependent variable is still present in each equation.

25

Because each version of the model has similar standard errors, only the results for the first version—OLS, equation (20′)—are reported here.

26

In the case of tin, the level of inventory is significant at 5 per cent for the INV estimation of equation (20″) (See the equation for tin in Table 5.), although it is not as significant in the other three estimations.

27

This is not to say that the flow variables are not important in determining price; on the contrary, they are important.

28

The constrained version—that is, equation (20″)—has smaller standard errors for cocoa, coffee, and sugar in the OLS results; and, in addition to these three commodities, it also has smaller standard errors for cotton in the INV results.

29

The elasticities e are evaluated at sample means—that is, e = ∂Pt/∂Pt+1eP¯t/P¯t+1e. Therefore, e = ∂Pt/∂PMtP¯t/PM¯t because Pt+1e = α0 + α1 PMt (See equation (16″).). Pt and PMt are, respectively, the sample means of Pt and PMt. The elasticity estimates are based on constrained INV estimates (See Table 5.). As was noted earlier, the estimated long-run elasticity of the tin price with respect to expected prices was constrained to have a value of unity.

30

The coefficient of Pt–1 is not significantly different from zero, which implies that the speed of price adjustment r approaches infinity.

31

Cooper and Lawrence (1975) have made a thorough study of the commodity boom.

32

The tin equation was excluded from this exercise because a structural shift that occurred in 1974 was detected during the process of estimation.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Papers
IMF Staff papers: Volume 26 No. 1
Author: International Monetary Fund. Research Dept.
Volume/Issue:
Volume 1979: Issue 001
Publisher:
International Monetary Fund
ISBN:
9781451956528
ISSN:
1020-7635
Pages:
232
DOI:
https://doi.org/10.5089/9781451956528.024

World Cocoa and Coffee Prices, 1955–75

(1970=100)

World Cotton Prices, 1963–75, and World Rubber Prices, 1955–75

(1970 = 100)

World Copper and Tin Prices, 1956–75

(1970 = 100)

World Sugar Prices, 1956–75

(1970 = 100)